1. Field of the Invention
The present invention relates to a color processing apparatus and method that perform color reproduction that corresponds with the brightness of the subject at the time of photographing.
2. Description of the Related Art
Generally, since preferable color reproduction differs depending on the photographing scene (brightness of subject at time of photographing), it is necessary to adjust the color reproduction in accordance with the photographing scene. As technology for reproducing colors in accordance with the photographing scene, technology exists that reproduces colors so as to impressively match the appearance at the time a photographer photographs a subject on a print or a monitor on the viewing environment side. This technology uses a color appearance model (CAM) such as CIECAM97s or CIECAM02 that are recommended by CIE (for example, see Japanese Patent Laid-Open No. 2006-080907).
The aforementioned color appearance model will now be briefly described. CIECAM02 that is one example of a color appearance model uses mathematical expressions to model the degree of adaptation with respect to an illuminating light or to model visual phenomena such as a contrast effect produced by the background. More specifically, based on tristimulus values XYZ of a given color, CIECAM02 transforms the color into a perceptual color space that is expressed by attributes (perceptual brightness J value, perceptual saturation C value, perceptual hue h value) that represent the appearance of a given color as the human perception by taking into consideration the viewing conditions of the visual environment. Utilizing this color appearance model, viewing environment parameters are set for the input side and the reproduction destination side, respectively, and by performing transformation from the input side and inverse transformation to the reproduction destination side so that values on the perceptual color space match, it is possible to obtain colors having the same appearance even though the tristimulus values thereof differ.
In recent years, a color appearance model adapted for still images (iCAM) has been proposed by the CIE. In iCAM, local chromatic adaptation processing and visual frequency characteristics are taken into account to expand the color processing of CAM. As the local chromatic adaptation processing of iCAM, the field of view of the viewer is defined, viewing environment parameters are calculated based on the pixel data distributed in the field of view, and viewing environment parameters are set for each pixel in the image to perform color transformation.
As an example of viewing environment parameters of a color appearance model, FIG. 13 shows viewing environment parameters for CIECAM02 as well as the recommended value for each parameter.
However, the following problem exists with the above-described conventional color reproduction technology that uses a color appearance model. That is, with the viewing environment parameters recommended by CIECAM02, because all pixels of an image undergo color transformation using the same viewing environment parameters, there are cases in which the appearance cannot necessarily be appropriately reproduced. For example, as the luminance of the adapting field that is one viewing environment parameter, gray corresponding to a luminance value of 20% of white luminance is set that is assumed to represent the average brightness of a subject. Accordingly, optimal color transformation is not executed on pixels in regions in which the brightness is higher or lower than the gray in question in a subject image.
Further, when using iCAM as a color appearance model, since color transformation is performed by setting viewing environment parameters for each pixel, there is the problem that although the appearance of the defined field of view is favorably reproduced, in some cases the color balance when the entire image is viewed is not favorably reproduced. Particularly, in a case in which the defined field of view is narrow, since the adapting white point approaches the chromaticity point of the pixel value of interest, the saturation of the overall image decreases.